I would guess it assumes the "bearing" doesn't crush or move?
You say the force concentration shows up at opposite end of crank from which the force is applied? Is this a steady state (analyzes one frame in time) or dynamic analysis (takes into account the actual movement & "records" some parameter(s) like stress, strain, deflection, movement, twist, etc? Is force concentration on oppopposite side of crank as applied load or same? i.e. crank is bowing (forces on same sides) or teetering (forces on opposing sides) about some axis?
Simulated bearing responds like a real bearing with a support factor. So there is some radial movement, but it is rated in pounds per inch displacement. Kinda like a very very stiff spring. Which is an accurate model of a real bearing.
This is steady state right now. You have to develop the states before you can move on to the dynamic and harmonic simulations.